基于灰关联和神经网络的通风系统评价模型  

Mine Ventilation System Assessment Model Based on ANN and Grey Correlation

作　　者：刘庆龙[1] 魏夕合[1] 史俊伟[2] 吕康[3] 

机构地区：[1]山东科技大学资源与环境工程学院,山东青岛市266510 [2]山东工商学院管理科学与工程学院,山东烟台市264005 [3]中国矿业大学(北京)资源与安全工程学院,北京100083

出　　处：《矿业研究与开发》2012年第2期63-66,共4页Mining Research and Development

基　　金：国家自然科学基金项目(51074099);山东科技大学研究生科技创新基金项目(YCA110303)

摘　　要：针对BP神经网络随维数增加学习时间剧增,易陷入局部最小点以及通风系统评价样本量小、评价指标多的技术难题,利用灰色关联度筛选评价指标,以此确定人工神经网络的输入参数。采用基于Levenberg-Marquardt(L-M)算法的BP神经网络,建立通风系统综合集成评价模型,并用VC++6.0开发了灰关联—神经网络评价软件。选用基于人工神经网络通风系统评价的典型实例作为评价样本,将原有16-32-3的网络模型简化为8-3-3。结果表明,建立的评价模型仿真结论与基于人工神经网络的结论完全吻合,并且模型简单,易于操作,训练效率大幅提高,有一定的推广应用价值。In view of learning time of BP neural network increasing sharply with dimension,network easily falling into the local minimum point,and a little of assessment sample and many assessment indexes for ventilation system,grey correlation was applied to select assessment indexes,so as to determine input parameters of artificial neural network.BP neural network based on Levenberg-Marquardt algorithm was used to establish a meta-synthesis assessment model of mine ventilation system,and the software for grey correlation-neural network evaluation was developed using VC＋＋ 6.0.Taking a ventilation system assessment example based on artificial neural network as assessment sample,16-32-3 network model was simplified to 8-3-3 network model.The result showed that the simulation results obtained by the established assessment model were completely consistent with the simulation results based on artificial neural network,and the model is simple and easy to operate,the training efficiency of the model is raised greatly,so the model has a value to be applied in certain extent.

关 键 词：通风系统 评价模型 灰色关联 神经网络 LEVENBERG-MARQUARDT 

分 类 号：TD724[矿业工程—矿井通风与安全]